Research On Channel Estimation Based On Pilot In MIMO-OFDM Ystems

The channel estimation techniques in wireless communication systems has been the research focus, in order to ensure the correct demodulated data accurate channel estimation is necessary. To meet people’s high speed, high quality and communication needs of multi-service is important. This paper mainly studies estimation and channel estimation based on pilot performance improvement channel.This paper introduces the basic principle of OFDM systems, Jakes channel model, pilot channel estimation algorithms and the interpolation algorithms and in this paper by MATLAB simulation of the performance of OFDM system different channel estimation and interpolation algorithms. Introduces the MIMO-OFDM system model and STBC coding, analyzed multi-antenna LS channel estimation algorithm and optimal pilot sequences property.And simulation of the MIMO-OFDM system performance using different frequency insertion channel estimation algorithm mean square error (MSE) and bit error rate (BER) performance.Based on DFT noise reduction algorithm truncated noisy to filtered noise. In this peper, after the analysis of existing literature,we put forward two of DFT noise reduction improved algorithm---DFT noise reduction algorithm based on mean shift and DFT noise reduction algorithm based on dynamic threshold. DFT noise reduction algorithm based on mean shift by selecting offset to filter out more noise, but the offset selection is fixed. DFT noise reduction algorithm based on dynamic threshold Mahalanobis distance as a criterion to select the threshold, the threshold can be changed with the noise power. Both algorithms to overcome the disadvantages of the prior literature suggests that algorithms can not filter out more noise by setting a threshold is reached with a drop more noise. Thesis of these two algorithms with existing algorithms for channel estimation and BER performance simulated and compared. The simulation results show that the two algorithms can improve channel estimation and BER performance, including dynamic threshold based on the best performance of DFT noise reduction algorithm.Traditional channel estimation algorithm considered is a wholly-intensive channel, channel estimation need to be inserted more pilots to reconstruct the entire channel impulse response. Compressed sensing channel estimation algorithm can reduce the pilot insertion and raise spectrum efficiency. In this paper, compressed sensing channel estimation algorithm were studied, we analyzed the compressive sensing channel estimation process and use of the algorithm for the simulation. Simulation results show that the compressed sensing channel algorithm can use fewer pilots to get a good estimate of performance.